In summary, the narcolepsy-cataplexy patients rated the cartoons as being less humorous than the healthy control group. Results from the brain imaging analysis showed that although both groups activated regions associated with humor appreciation, the cataplexy group showed significantly more activity in the nucleus accumbens and hypothalamus as well as the right inferior frontal gyri when directly compared to the healthy control group. Intriguingly, examination of the one patient experiencing the cataplectic attack showed dramatic reductions in hypothalamic activity. Despite the paradoxical nature of this finding (between the cataplexy group and the single cataplectic attack patient), one could suggest that hyperactivation of certain parts of the emotion circuitry lead to subsequent shutdown of the hypothalamus resulting in the characteristic REM-sleep like state.
It is know that the limbic network, including the ventral striatum, hypothalamus, and amygdala, show increased activation during laughter and rapid-eye-movement (REM) sleep (e.g. 
). It is also notable that the hypothalamus, where hypocretin neurons are located, is also likely involved in the generation of laughter as gelastic seizures are commonly observed with hypothalamic hamartomas 
. Thus, it is of particular interest that patients with cataplexy showed greater activation in these regions relative to healthy controls with humor stimulation. We have previously observed that the degree of humor rated by healthy subjects is positively correlated with activation of these neural systems 
. In this respect, the observation of greater activation in subjects with narcolepsy-cataplexy is paradoxical in light of the relatively lower humor ratings by this group. One explanation for this is that the lack of hypocretin in patients with cataplexy intensifies humor-induced activation of these regions, possibly exacerbating normal physiological reactions to humor.
Intriguingly, results from single photon emission computed tomography (SPECT) studies of cataplexy have shown abnormal perfusion in the hypothalamus and basal ganglia. For example, Chabas et al. 
, found restricted hyperperfusion of the cingulate, orbitofrontal and right temporal cortices, and putamen in comparison to a baseline, non-cataplectic state in affected individuals. Hong et al. 
used a similar technique in two cataplexy patients and found hyperperfusion of the amygdala and basal ganglia (as in 
), but also hypoperfusion of prefrontal cortex during cataplectic attack. While these results are restricted by low sample size and poor spatial and temporal resolution of SPECT, they suggest that cataplexy is produced by hyperactivation and hypoactivation of amygdalo-cortico-basal ganglia–brainstem circuit.
Our results are in partial support of a recent fMRI study of humor in narcolepsy-cataplexy patients 
. In that study, and similar to the results presented here, participants with narcolepsy-cataplexy showed heightened activity of the amygdala and nucleus accumbens when exposed to humorous stimuli. Unlike controls however, these subjects did not show altered activity in the hypothalamus in response to humorous (or non-humorous) trials. Abnormal function of the hypothalamic-amygdalar circuit was proposed as one route from which positive emotions may trigger catapletic attacks 
. Findings in our cataplectic attack patient support this contention, and further suggest that massive suppression of hypothalamic activity may an essential component of a cascade of neural events leading to muscle atonia. Collectively, these fMRI studies of cataplexy indicate that full blown attacks result from an inability to control the emotion system, including the hypothalamus.
The fact that the narcolepsy-cataplexy group showed relatively greater activation of right inferior frontal gyrus, a region known to be involved in inhibitory control 
is important in the context of the behavioral results. The right inferior frontal gyrus receives direct input from the striatum 
and is part of the striatal-thalamocortical loop. Anecdotal observations support the notion that cataplexy patients train themselves to suppress laughter and avoid humorous material. This suggests that the right inferior frontal gyrus is involved in the conscious effort to prevent the expression of humor associated behaviors 
and keeping emotional material out of mind 
. Decreased activation of the prefrontal cortex during status cataplecticus has also been reported 
. The right inferior frontal gyrus also modulates the emotional network including the nucleus accumbens and hypothalamus 
. In this context, it is tempting to speculate that the activation of one or more components of the inhibition “network” serves to prevent exacerbation of normal physiological reactions to humor such as transient muscle weakness reported by healthy subjects when laughing 
, and that patients with cataplexy further activate this compensatory mechanism to prevent cataplexy.
In conclusion, we demonstrate that patients with narcolepsy-cataplexy show overactivation of the nucleus accumbens and hypothalamus. In addition, these subjects also showed increased activation of the inhibitory network which might account for lower ratings of humorous stimuli and suppression of laughter. These findings therefore suggest that increased activation of the nucleus accumbens and hypothalamus might contribute to the symptoms (e.g. muscle atonia) observed in this population, while the inhibition network acts to compensate and may represent top-down cortical control of cataplectic symptoms. Whether severe cataplexy symptoms are marked by a “overactivation” and “shut-down” of the hypothalamus remains an important question for future research.